Following on his previous talks on data visualization and programming interfaces, Bret Victor presents the idea of what he calls a “seeing space”, meant to improve understanding of problems in the context of collaborative engineering.
Here are a couple of links on how to render light scattering effect (aka. volumetric shadows):
- GPU Gems 3, chapter 13: Volumetric Light Scattering as a Post-Process
- SIGGRAPH 2010 paper: Epipolar sampling for shadows and crepuscular rays in participating media with single scattering
- Intel article: IVB Atmospheric Light Scattering
- Unity plugin implementing above Intel article: Light shafts Unity plugin
- Slides: Volumetric Lighting for Many Lights in Lords of the Fallen
- Sébastien Hillaire wrote an article on dynamic volumetric lighting with temporal upsampling.
In this (slightly over) one hour talk, 1½ hour including Q&A, John Carmack walks through the physics of light, the early days of rendering, the current state of the art, and the direction it is headed at. In short: until we can afford path tracing, we’re approximating it.
Last month at SIGGRAPH, Michał Iwanicki of Naughty Dogs presented his talk “Lighting technology in The Last of Us”, in which he focused on the technique they used for ambient shadows. In short: light maps and analytic occlusion with ellipsoid approximations of objects. Clever!
Quaternion are a very useful tool in 3D, but also one that is unintuitive and difficult to get a natural feeling about. The talk Jim Van Verth, of EssentialMath, gave earlier at GDC2013 explains some facts about quaternions and how they work, by looking back at their discovery: Understand Quaternions.
Update: on a side note, here is a trick for faster quaternion – vector multiplication.
This is the third day at Revision, and my contribution this year is the talk I gave yesterday. Unlike last year, this seminar is not technical at all but focused on the design aspect and, to some extent, how it relates to the technical one. The context is demomaking, but many ideas are still valid in other media.
There were some issues with the recording unfortunately, which means some elements are missing (you will notice some blanks at the beginning). In particular after 5mn, there is an important point which was completely cut out. The text was:
Throwing a new technique at whatever you’re doing is not going to make it any better. It’s only going to change what you can achieve. There are two sides of image creation: the technical one and the artistic one. Different techniques allow to do different things, and the more techniques you master, the better you understand what you can and cannot do with them, and how to do it. Technique becomes a tool that changes how you can express yourself.
For more demoscene related talks, here is the full list of seminars at Revision 2013.
Geeks3D mentioned this keynote at the GPU Technology Conference, where NVIDIA’s CEO shows their technology called FaceWorks. After talking about the uncanny valley and avatar rendering uses, came the live demonstration, which seriously raises the bar in terms of face rendering and animation.
The quality is incredible, the gap from photo realism is getting very narrow, and some expressions are really convincing. The transitions and frozen expressions feel weird though, so I am wondering how it would look running freely for a moment, with all the rapid subtle moves we show even when staying idle. The avatar as a mean of communication is certainly very appealing. It would be interesting to see if when facing this rendering, we would react to the expressions displayed.
The demo itself starts after 8 minutes.